Pack assembly and kit inspection
Confirm presence or absence of items in medical kits
Graphical programming environment for deep learning-based industrial image analysis
Kitting or system and procedure pack applications must verify that the correct item is in the correct location of a medical or vaccine kit and confirm the presence or absence of items. Regulations also require patient implant cards, directions, and/or caution documents to be included in the assembly. Medical device manufacturers typically rely on slow and expensive visual inspections performed by human operators or rule-based machine vision systems. Visual variations of products can cause complications for manual and automated inspection systems where it must be able to identify, count, and classify different items with varying package orientations, reflective surfaces, and complex geometries.
Cognex vision sensors perform a simple count of the objects in a package, passing or failing packages based on the programmed value. They verify that all medical devices and other items are present even under shrink-wrapped packaging, helping manufacturers error-proof their operations and prevent faulty packs from entering the supply chain.
Deep learning systems automate the task of locating, identifying, and classifying multiple features within a single image. It generalizes the distinguishing features of various items based on their size, shape, and surface features. A user can train the part location and assembly verification tool to locate each type of item that needs to be found. Once trained, the image can be split into different regions where the tool will check for the presence of an item as well as verify that it is of the correct type. Deep learning also reliably locates and identifies inserts in boxes regardless of orientation and lighting conditions to prevent recalls and ensure patient safety.